GitHub Repo Briefing
OpenMontage
An agentic video-production system that coordinates media preparation, planning, editing, and rendering through a pipeline called Backlot.
A first test
Run make setup and open one sample project with python -m backlot open. Use test footage, inspect each pipeline step, and verify the final render before adding client media.
codebase-memory-mcp
A local MCP server that builds code intelligence from Tree-sitter parsing and a graph, with a static binary and support for many programming languages.
A first test
Install the binary and index a small repository. Ask for a symbol definition and callers, compare both with the source, and check the MCP client permissions before enabling edits.
no-mistakes
A Git command workflow that runs review, tests, documentation, lint, and push checks in a disposable worktree before it creates a pull request.
A first test
Run git push no-mistakes on a small test branch. Read every check and failure, then verify that the worktree and pull request contain only the intended files.
design.md
A design-document format that combines YAML design tokens with Markdown rationale so coding agents can use a visual system and its decisions together.
A first test
Create a small DESIGN.md for one screen with three tokens and one component rule. Ask an agent to implement it, then compare the CSS values with the declared tokens.
daily_stock_analysis
ZhuLinsen/daily_stock_analysis
A multi-market stock-analysis application that combines market data, news, an LLM analysis workflow, dashboards, and optional notifications.
A first test
Run python main.py --dry-run with a public symbol and inspect the input date and sources. Compare the output with the raw data; do not place trades from the generated analysis.
penpot
An open-source design platform for interfaces, components, variants, and design tokens, with a self-hosted deployment option.
A first test
Start the documented self-hosted demo and rebuild one small component. Check token inheritance, component variants, and export output before migrating a real design file.
Agent-Reach
A command-line skill that gives agents access to web pages and services such as X, Reddit, YouTube, GitHub, Bilibili, and Xiaohongshu through fallback backends.
A first test
Install it and fetch one public post from one supported service. Check the returned URL, timestamp, and text against the source before using it in a report.
hiring-agent
A Python application that evaluates and scores resumes with an AI model. The repository describes it as a resume evaluator, not an applicant-tracking system.
A first test
Run it with synthetic resumes and a local model. Compare the score with a human rubric, inspect the extracted evidence, and remove personal data before any real hiring test.
ai-website-cloner-template
JCodesMore/ai-website-cloner-template
A Next.js template that gives an agent a website URL and uses browser access to recreate the page structure and styling in a local application.
A first test
Run /clone-website on a site you own or have permission to copy. Compare the generated page with the source, replace protected assets, and check mobile layout before reuse.
voicebox
A local-first voice studio for text to speech, voice cloning, dictation, and MCP voice tools. It keeps the models and audio workflow on the user's machine when configured locally.
A first test
Start with a short script and one voice model. Generate a ten-second clip, check the output file and pronunciation, and confirm where the audio and model files are stored.
orca
A desktop and mobile workspace for running several coding agents in parallel, with separate worktrees and support for agents such as Codex, Claude, OpenCode, and Pi.
A first test
Run two agents on separate worktrees for the same small issue. Compare their diffs and merge one manually; check that neither agent can modify the other's branch.
agent-native
A framework for defining application actions that can be used from a user interface, agent, HTTP, MCP, A2A, or CLI client.
A first test
Define one read-only action and call it from two supported clients. Assert the same input validation and result shape before adding an action that changes data.